| name | goaly-screen-lead |
| description | Use when an inbound lead needs vetting before [Owner] engages. Deep research on profile, company, flags, fit assessment, and optional reply draft. |
Screen Lead
Before You Start
- Read
.claude/skills/_shared/conventions.md
- Read
gotchas.md in this directory
Context
- Today: !
date +%Y-%m-%d' ('%A')'
- Uncommitted changes: !
git status --short notion-mirror/
Learned Preferences
!tail -20 "${CLAUDE_SKILL_DIR}/data/feedback.log" 2>/dev/null || echo "No feedback yet — preferences will accumulate over sessions."
Trigger Phrases
/goaly-screen-lead or /goaly-screen-lead [name or email]- "screen this lead", "research this lead", "check this person out"
- "new lead from [source]"
Input
Accept any combination of: Name, Email, Company name, Website URL, LinkedIn URL, raw lead form submission, or "check emails for [name]" (triggers Gmail search first).
Quick Start
/goaly-screen-lead Contact-9, Tripadmit
/goaly-screen-lead contact1@example.com
/goaly-screen-lead "check emails for Contact-8"
Process
GATHER → ANALYZE → ACT (optional, only after [Owner] confirms interest)
Phase 1: GATHER
Collect all available information. Run these in PARALLEL where possible.
1. Email search (if email reference given):
gog gmail search '[name] newer_than:30d' --max 10 --account you@example.com --plain
2. LinkedIn research — Use perplexity_search or web_search_exa. Check: career history, title consistency across platforms, connections, time at company, skills.
3. Company research — Try curl -s -H "Accept: text/markdown" [url] first, fall back to crawling_exa. Check: what the business does (strip buzzwords), stage, revenue signals, funding, real content vs stock.
4. Company registry — UK: Companies House API. Ireland: CRO. US: State SOS / SEC EDGAR. Note if no verifiable entity found.
5. Web presence — Search "[name] [company]" via perplexity_search or web_search_exa. Find press, speaking, community involvement, social media (sherlock "[name]" if warranted), published content.
6. Tech signals — Visible stack (job postings, GitHub, BuiltWith). Flag proprietary/unknown platforms with no docs or community.
7. How they found [Owner] — Cross-reference language with [Owner]'s blog/site. Content-sourced leads are warm but may have inflated expectations.
Phase 2: ANALYZE
Compile findings into a structured report. Present directly in chat:
## [Name] — Lead Research Report
### Profile Summary
[2-3 sentences: who they are, career background, credibility level]
### Company Summary
[What it does, stage, revenue signals, team size, funding status]
### Green Flags
- [Specific budget, already running, phased approach, verifiable career,
content-sourced, retainer mention, Clear problem, team in place]
### Red Flags
- [AI buzzword overload, vague/stock website, Title shifting,
proprietary unknown tech, ASAP + ambitious scope,
No budget mentioned, pre-revenue, decision-maker absent,
wants free prototype, can't articulate without buzzwords,
expects blockchain/crypto build]
### What They Actually Want
[Strip ALL buzzwords. 3-5 bullet points in plain language.]
### Fit for [Owner]
**Alignment:** [What matches [Owner]'s services]
**Concerns:** [Budget risk, scope risk, timeline risk, payment risk]
**Engagement model:** [fCTO retainer, web dev retainer, advisory, pass]
**Bottom line:** [1-2 sentence recommendation]
### Qualifying Questions
[3-5 key questions tailored to what's still unknown]
See memory/lead-screening.md for detailed red/green flag definitions and screening checklist.
Phase 3: ACT (only when [Owner] confirms interest)
Do NOT proceed here until [Owner] says "interested" or explicitly asks.
1. Create client files — Write notion-mirror/clients/[slug].md and notion-mirror/contacts/[name].md with proper frontmatter. Check for duplicates first (see gotchas.md).
2. Draft reply — Read clients/tone-of-voice.md first. Rules:
- NEVER offer or suggest calls (never offer a call in any draft) — email only
- Angle toward retainers, not fixed-scope projects
- Be upfront about boundaries (what [Owner] doesn't do)
- Ask budget question directly: "Is that budget confirmed and ready to go?"
- First email qualifies, doesn't promise
- Present draft to [Owner] for review before sending
3. Create Notion task — In notion-mirror/tasks/:
- Project: Business Development (
YOUR-PROJECT-ID-4)
- Goal: Build Portfolio of Internet Companies (
YOUR-PAGE-ID-PORTFOLIO-GOAL)
- Area: Sales | Timeframe: This Week | Energy: Research | Impact: Supporting
4. Update MEMORY.md — Add new lead to Active Leads section.
Session Learning
When [Owner] corrects your output or expresses a preference during this session, immediately append it to the feedback log:
mkdir -p "${CLAUDE_SKILL_DIR}/data" && echo "$(date +%Y-%m-%d) <preference description>" >> "${CLAUDE_SKILL_DIR}/data/feedback.log"
Only log general preferences that apply to future sessions — skip task-specific corrections. Use judgment on detail: some preferences are one line, others need a sentence or two of context to be useful in future sessions.
Error Handling
| Scenario | Action |
|---|
| No LinkedIn found | Yellow flag, search alternate spellings |
| Company website down | Check Wayback Machine, note as flag |
| No company registry record | Flag unregistered entity, check alternate jurisdictions |
| Vague lead brief | List gaps, suggest qualifying questions |
| Warm intro lead | Note referrer and relationship — adjusts posture |
| No email in Gmail | Ask [Owner] for lead source, web research only |
| Multiple people with same name | Present options, ask [Owner] to confirm |
Output
Present the structured report in chat. Then ask [Owner]:
- Interested — Create client files, draft reply, create task, update MEMORY.md
- Maybe — Create screening task with "This Month" timeframe
- Pass — No action needed
Related Files
| Purpose | Path |
|---|
| Screening checklist & flags | .claude-memory/lead-screening.md |
| Tone of voice guide | clients/tone-of-voice.md |
| [Owner]'s profile & preferences | .claude-memory/dan-profile.md |
| Revenue strategy | .claude-memory/revenue-strategy.md |
| Research tools reference | docs/claude/research-tools.md |
| Notion reference data | docs/claude/notion-reference.md |
| Contract templates | clients/templates/ |
